论文部分内容阅读
在卫生政策的效果评价中,研究的主要目的就是推断卫生政策与政策干预效果之间的因果关系。但是,卫生政策研究无法根据研究者的意愿将研究对象随机分组,一些具有较高因果关系解释力度的研究设计在实际应用中往往会受到限制,存在一些因素可能诱导我们对解释变量和效果变量之间的因果关系做出错误推断。由此产生的遗漏变量偏倚、选择性偏倚以及信息偏倚通常是通过产生所谓的“内生变量”问题而影响研究结果的。工具变量模型正是一个能够有效解决观察性研究中内生变量问题的统计方法。所以,本研究结合了卫生政策效果评价研究的特点,分析工具变量模型在该研究领域中的应用及前景,为解决卫生领域研究中内生变量问题提供一个新的思路。
In the evaluation of the effect of health policy, the main purpose of the study is to infer the causal relationship between the effectiveness of health policy and policy intervention. However, the health policy research can not randomly divide the research subjects according to the wishes of the researchers. Some research designs with high causal explanatory power tend to be limited in practical application. There are some factors that may induce us to interpret the explanatory variables and the effect variables The causal relationship between make mistaken inference. The resulting missing variables bias, selectivity bias, and information bias often affect research outcomes by creating so-called “endogenous variables.” The instrumental variable model is a statistical method that effectively solves the problem of endogenous variables in observational studies. Therefore, this study combines the characteristics of the health policy evaluation study, analyzes the application and prospect of the tool variable model in the research field, and provides a new idea for solving the endogenous variables in the field of health research.